New Tool From Drawbridge Helps Marketers Bridge the Gap Between Desktop and Mobile

Sometimes it’s better to listen. For Drawbridge, one of the hottest startups in advertising technology, it’s a lesson that’s spawned a business, and a new product out today that helps marketers retarget users across platforms and devices in a move to help increase retention.

The company has spent the last two years analyzing the billions of ad requests that come through mobile and desktop ad exchanges everyday, developing a database that matches multiple devices based on their user and subsequently allows marketers to break down the barrier between desktop and mobile advertising. Now, seven months after bringing its two flagship products to market — a cross-device marketing tool and a mobile audience retargeting tool — the startup has launched a new mobile retargeting product that enables marketers to target existing users based on a previous activity — say, an app download — as they move across apps, mobile web, and desktop browsers.

“There’s an iron clad barrier between the mobile app and the mobile web worlds today. If I knew that there were two ad requests coming from a mobile application and from a mobile web browser, each of which came from the same device, I would have no way of knowing that it came from the same user,” Kamakshi Sivaramakrishnan, the company’s founder and CEO, told Street Fight in an interview. “The way we solve for that problem is to use an inferred identity to bridge that gap.”

By breaking down the barriers between mobile and desktop, the data scarcity problem in mobile quickly fades, and the the value proposition improves.

That inferred identity — the backbone of Drawbridge’s technology — comes from the company’s machine learning algorithms that find patterns in the bits of data that publishers include in ad requests. And one of the most valuable data points, says Sivaramakrishnan, who holds a Ph.D in information theory from Stanford, is location. By listening to the location data, generated from a user’s IP address and included in most mobile impressions, these algorithms can begin to see which devices live, travel, and work in the same places.

As more information is passed through the algorithms, their ability to match these device improves dramatically. The numbers vary based on user activity, but Sivaramakrishnan says it typically takes three to four weeks to achieve a 60% confidence that two devices belong to the same person. Overall, the company has mean confidence rate of 60-70% for the half of a billion matched devices on its platform.

In November 2012, the company launched its first two services on top of the match technology, a mobile and cross-screen marketing product, both of which use its matching service to allow marketers to use data generated on one device — say, a visit to a travel booking site — to interact and evaluate a user’s behavior on another device, like a mobile phone. For instance, if a user clicks on a mobile ad for Macy’s on the web, then uses the company’s mobile app to look up a store nearby on the way home from work, Drawbridge can connect the dots.

But both products focused on customer acquisition. The new mobile-to-mobile retargeting product, which the company launched today, looks to bring some of that retargeting technology to a business’ existing user base, providing a middle ground of sorts for mobile reengagement. Instead of relying on deep, and intrusive, engagement tools like push notifications on a mobile apps, the product allows marketers to channel those notifications into paid media ad formats.

“Once a brand or developer has a user, there’s a spectrum of engagement… from highly engaged to entirely inactive,” says Sivaramakrishnan. “We’re allowing brands to engage a certain segment of these users — maybe those who have lapsed for a month — across the [mobile and desktop web.]”

The technology, which Drawbridge has pioneered, could deeply shift the wiring of the mobile advertising industry. By breaking down the barriers between mobile and desktop, the data scarcity problem in mobile quickly fades, and the the value proposition for companies focused on solving that issue becomes substantially less acute.

Within the hyperlocal space, that could spell trouble for a number of mobile-local adtech startups that have invested heavily in building technology that use location to replicate the audience profiles widely used on the desktop web. If the barrier between desktop and mobile data is dissolved, these companies now will compete with the established cookie-based audience profiling firms that have scaled on desktop over the past few years.

However, it also presents a potentially lucrative scenario for marketers: Use the data generated on the web to identify an audience (say, sushi lovers) and then leverage data generated on mobile — namely, location — to understand context (on the way home from work). In many ways, the lumping of those two concepts — audience (who we are) and context (what we’re doing) — is a vestige of a desktop environment in which our context was static, limited to our homes, office, and possibly a coffee shop. Context, not audience, is the big mobile problem; and it’s still not close to being solved.

Steven Jacobs is Street Fight’s deputy editor.